To the Editor:
The paper by W. Scott Bennett (“Normalized Site Attenuation and Test Site Validation”, Summer Issue No. 178) suggested that the vertical polarization normalized site attenuation (NSA) values in ANSI C63.4-1992 are in error by as much as 1.5 dB. The author failed to acknowledge that the NSA values in C63.4 are based on the radiation pattern of an ideal point dipole, whereas his calculations assume an isotropic radiation pattern.
Zhong Chen and Tim Harrington
EMC Test Systems
(Ed. Note: Additional comments by Zhong Chen and Tim Harrington are included in this issue’s “Practical Papers, Articles and Application Notes” column.)
To the Editor:
The article by Bennett on “Normalized Site Attenuation and Test Site Validation" in the Summer issue of the IEEE EMC Society Newsletter contains a major error: He has neglected the angular dependence of the radiated field from a vertical electric dipole. He assumes that the fields at the Observation Point in his Fig. 1 are isotropic in the plane of the drawing.
While this is true for a horizontal electric dipole source, it is certainly not true for a vertical electric dipole. I think he will find complete agreement with the ANSI C63.4 values if he includes the appropiate term, proportional to the sine of the angle between the direction of the dipole and the direction of propagation.
There are some errors in the numbers given in C63.4 caused by neglecting the near field terms (which Bennett also neglects), but these are appreciable only near 30 MHz, as I point out in “Near-Field Corrections to Site Attenuation”, IEEE Transactions on EMC, vol. 36, pp. 213-220, 1994.
David Gavenda
Physics Dept
University of Texas at Austin
To the Editor:
Normalized site attenuation (NSA) is defined in ANSI C63.4-1992 as follows:
Site attenuation divided by the antenna factors of the radiating and receiving antennas (all in linear units)
That is precisely equivalent to assuming uniformly radiating and receiving antennas both of which are free from near-field effects. And, that is precisely what was done to obtain the results given in my article.
W. Scott Bennett
Carr, Colorado
To the Editor:
I’m guilty, . . . you’re guilty, . . . we’re all guilty — guilty of not giving enough thought to the things we do. Many years ago, when I worked as a technician for IBM, everywhere you looked there were one-word signs that said “THINK!” Whether they still have those signs today, I do not know. However, it becomes clearer with every day that we need those THINK signs now more than we ever did. And, we need them everywhere! One recent example of that need is the failure of almost everybody’s computers to distinguish the year 2000 from the year 1900. It is quite apparent to all of us, now, that “- - 00" could represent either 1900, or 2000 (or any of 98 other numbers, for that matter!). However, a few appropriately placed THINK signs might have made that apparent to computer software designers a lot earlier, and averted the problem.
Another example of our need to think has bothered me for many years. How often have you heard, or read (or said, or written) the words “current flow”? Those words are used over and over by almost all of us, even in well-respected books and journals. However, the words are seldom used correctly! Current, i(t), is charge flow, dq(t)/dt. Current flow is the propagation, or acceleration, of current, di(t)/dt. Without any exception I am aware of, however, when someone says “a current flows”, they mean “a current exists”! So, their words and the meaning intended are not the same. We all need to be much more precise. Otherwise, the words “teach”, “learn”, and “understand” will not have the meanings intended either.
More specific to EMC . . . a recent paper of mine on the radiations of small loops of current won third prize for best paper at an annual IEEE EMC Conference. However, shortly after rejoicing I realized that my analysis was incomplete and my prize-winning paper was incorrect! I had overlooked the reciprocal effects of their fields on the currents — something that should never be overlooked! Nevertheless, only a few people thought I was wrong, and not one said that a neglect of reciprocity was the reason. Typical of the reasons given was that some highly abstract mathematical equation was not satisfied. This example clearly shows that fundamentals are not getting the attention they should get, from any of us.
Two more recent papers of mine point a finger at every one of us who has worked in EMC for the last decade or so. Some of the numbers given in ANSI C63.4 are easily seen to be wrong, but we have all accepted them for more than 10 years! The values in error are those given for normalized site attenuation (NSA) with vertical polarization. All of those values given in that standard are in error — some by only 0.1 dB, but several by as much as 1.4 dB! That is very easily inferred from the definition of NSA and from the perpendicularity of an E-field and its H-field. The 10-year life span of those errors in a national standard again clearly illustrates a widespread need to pay much more attention to basic principles.
From each of the above examples, it should be obvious that we all need to BACK UP and give much more thought to fundamentals. We all need to SLOW DOWN in using mathematics that cloud, even hide, basic relationships that can and should be made obvious. And, we all need to THINK of how to simplify and clarify the things we do, so that we can maximize our own understanding, and facilitate the sharing of it with others.
Sincerely,
W. Scott Bennett
Carr, Colorado